Study design was made as two groups pre-test, post-test experimental study where non probability convenient sampling technique was used for sampling. All 30 patients were explained about the study and informed consent was taken from them. After obtaining the ethical approval from the Ethical Committee (Ref: EC-FAC/18/PHY/072) thirty hemiplegic patients were allocated into two groups of 15 in each group. The study was conducted at OPD of Krupanidhi College of Physiotherapy and the duration of the study was from January 2018 to June 2018. The inclusion criteria were ischemic type of stroke, patients having Middle cerebral artery involvement, both male and female genders were included for the study. Criteria for inclusion were considered with the age limit 40 to 65 years. Patients under the category of Brunnstrom’s recovery stage 3 and 4 [24] were included. The duration of the condition from onset was less than six months and the patient should be able to walk 11 meters forwards by taking assistance or independently.

Hemorrhagic and traumatic with Posterior cerebral artery, Anterior cerebral artery involvement were excluded. Patients with age below 40 years and above 65 years and Brunnstrom’s recovery 1, 2, 5 and 6 were excluded for the study. Patients with cardio respiratory abnormalities perceptual and cognitive deficits and pre existing deformities of lower limb were also excluded. Brunnstrom’s stages of recovery indicates the severity of the stroke. It has six stages for lower limb: Stage 1: having flaccidity, Stage 2: having minimal voluntary movements, Stage 3: where the patient is able to flex hip, knee and ankle in sitting and standing, Stage 4: where the patient can attain sitting knee flexion beyond 90°, within the foot sliding backward on the floor, voluntary dorsiflexion of the ankle without lifting the foot off the floor, Stage 5: standing isolated non-weight bearing with knee flexion and hip extension and Stage 6: standing hip abduction beyond range obtained from elevation of pelvis, inner and outward rotations at the knee [24].

The treatment was one session per day, four days a week for four weeks. The study duration was for six months, Group A was provided with backward walking along with Conventional Physiotherapy and Group B was provided only Conventional Physiotherapy.

The experimental group (Group A) was given backward walking training along with conventional therapy. Patients were given instructions regarding backward walking training and demonstration was done. Kneeling at the side of the patient, the hemiplegic leg was moved in the correct pattern. With one hand, toes were held in dorsal extension and the other hand was placed over the patient’s buttock to prevent patient’s hip from hiking the pelvis upwards and backwards as the leg moves. Patient was instructed not to resist the movement of his leg. Total extension of leg was avoided by moving the foot back in small steps with knee flexion. To avoid plantar flexion of the spastic leg patient was instructed to keep the heel inwards during foot placement. When the patient had moved a step backwards with the hemiplegic foot, assistance was given to lower the heel to the ground as he had to take a step backwards with his sound leg. Treatment was given for 30 min with adequate rest periods in between depending on patient’s comfort.

Control group (Group B) was given only Conventional therapy. Tone normalisation was done by slow sustained stretching and cryotherapy for 10-15 minutes for the spastic agonist muscles. Weight bearing posture with slow rocking movements in quadruped position and Rhythmic rotations were given to the paretic limbs. Free ranges of motion exercises with voluntary movements were encouraged. Reaching forwards, sideways with Functional reach activities in sitting and standing were given as balance training. To retrain gait weight bearing exercises through legs, active flexion of leg maximally and a minimum flexion of leg to gain control of voluntary movements were encouraged; which is a preparation for walking. Proprioceptive Neuromuscular Facilitation (PNF) technique movement patterns, D1 extension with knee flexion for the paretic leg were given.

Stop Watch and measuring tape were the measurement tools. Patient was instructed to walk in his own pace and his gait parameters were assessed using Wisconsin gait scale. The distance of 11 meters was marked using measuring tape. By means of stop watch, the time taken to walk (9 meters) was measured in seconds. The first and last one meter was not included as the rate of acceleration and deceleration would vary, which in turn would alter the walking speed. The number of steps per minute (cadence) was measured by instructing the patient to walk in the straight corridor. For safety purpose patients were assisted by walking along with them.

Statistical analysis was done using independent ‘t’ test and paired ‘t’ test for between the groups and within the groups respectively. The data was analysed using the SPSS version 20.

After analysing the data the results obtained showed that there was significant improvement in Group A (Experimental group) on the measures of Wisconsin gait scale, walking speed and cadence, [Table/Fig-1]. There was also significant improvement in Wisconsin gait scale, walking speed and cadence in group B (Control group) [Table/Fig-2]. The results obtained after the analysis of independent ‘t’ test between the group showed that there was significant improvement in all three parameters in group A (Experimental group); Wisconsin gait scale (17.42±1.53), walking speed (14.4±2.11) and cadence (114.6±3.76), than Group B, Wisconsin gait scale (22.47±1.52), walking speed (20±0.5) and cadence (97.06±1.61) [Table/Fig-3]. Improvement was significant in the gait performance assessed using Wisconsin gait scale in experimental group with the score of 17.42 than the control group with the score of 22.47. Significant improvement was also seen in walking speed in the experimental group with the score of 14.4 than the control group with the score of 20. Improvement in cadence was seen in experimental group with the score of 114.6 steps per minute than 97.06 steps per minute of control group.

The person with hemiplegia demonstrates a slower gait pattern [9]. Recovery remains incomplete in most of the patients [14]. Backward walking has been proposed as a feasible and important tool in the gait rehabilitation [26]. Thus the study was aimed to examine the effects of backward walking training in stroke patients. The study result after treatment showed significant improvement in group A and group B in gait performance, walking speed and cadence. When compared between the groups, experimental group (group A) showed significant difference than the conventional therapy group (group B) in all three parameters: Gait performance evaluated using Wisconsin gait scale, walking speed measured marking 11 meters excluding first and last meter as the rate of acceleration and deceleration would alter the result and cadence, which is the number of steps in one minute. In a study carried out on 27 patients by Sang-Jin Kim et al., similar results were found in the improvement of walking speed, the average value increased from 68.12 cm/s to 92.60 cm/s in the ground training group and the cadence had improved from 73.09 steps/min to 85.36 steps/min among the 13 patients of ground training group [22]. Chang-Yong Kim had demonstrated in a study on 51 subjects, with three group of 17 in each group that lateral walking is better than backward walking training in hemiplegics with the mean value of 0.42 to 0.59 m/sec in lateral walking training group than 0.44 to 0.51 m/sec in backward training group in terms of walking speed, whereas it was 97.92 steps/min to 105.87 steps/min in lateral walking training group, better than 96.89 steps/min to 103.32 steps/min in backward walking training group in terms of cadence [27]. Although there was no significant difference between forward training and backward training among 10 cerebral palsy children assigned in two groups in a study conducted by Gazbare P et al., there was improvement in walking speed in both the groups with the intervention of backward walking training for 20 mins a day 3 days per week for 8 weeks in experimental group and forward walking training for the same duration in control group. They showed the mean value of 38.4 pre test score and 42 post test after 8 weeks of backward walking training, compared to 29.4 pretest score and 32.4 post test score showing significance difference in one minute walk test in both the groups, backward walking training group and forward walking training group, but no significant difference between the two groups p>0.05 [28]. Backward walking incorporates active hip and knee extension with ankle dorsiflexion which has the benefits which breaks the extensor pattern [21]. This could be the possible reason for the improvements in gait performance. Increased cortical activation may be an advantage of backward walking training by promoting engagement of intact cortical centers or descending drives to stimulate connections and activation of damaged spinal pathway [21]. In addition, the increased cerebral activation inherent in backward walking may have better engaged damaged cerebral circuits, leading to neuroplastic recovery that generalised to gains in forward walking ability [25]. As the patients were instructed to walk on self pacing speed, it could have influenced the patient’s cadence [22]. The improvements obtained in our study in the gait of stroke patients is similar to the study done by Deshmukh PD et al., which demonstrates that backward walking training is better than conventional therapy alone [29]. Studies suggest that backward walking training results in greater muscle and neural activities, higher heart rate, oxygen consumption and lower patellofemoral joint reaction forces [28]. This could be related to the number of steps increased after the sessions. Moreover backward walking had an increased oxyHb response in the supplementary motor area, pre-central gyrus and the superior parietal lobule. This suggests that walking backward had a greater burden on the medial sensorimotor cortices [30]. With the above discussed mechanisms it is proved that backward walking along with conventional therapy can result in greater walking ability in stroke patients.

The study was conducted in 30 subjects. In order to generalise these findings, it can be done with larger samples. The duration of the study was only four weeks. It can be carried out for longer duration to obtain better outcome. The follow-up was not studied after four weeks; long term effect of the intervention can be checked. The study can be done using modern digital gait analysers for better and accurate results.

Hemiplegic gait pattern has lesser values for speed, cadence and higher value for duration in the gait cycle. It was hypothesised that walking backwards is an effective approach in improving the walking pattern of hemiplegic patients. From the results obtained it is shown that, adding up backward walking in the rehabilitation protocol, there has been a significant improvements in the gait performance of the stroke patients. Therefore, it could be suggested that backward walking training may be an adjunct training to be incorporated with the conventional programme for improving patients’ gait performance in hemiplegia.